Biology Reference
In-Depth Information
neocortical responses. This mechanism explains experimentally obtained attentional
modulation of neural activity in the LGN, wherein neural responses to attended stimuli were
enhanced and responses to ignored stimuli were attenuated (Kastner and Pinsk, 2006;
O'Connor et al., 2002). Thus LGN may serve as a “gatekeeper” in attentional control of visual
responses.
It was shown that the cortical areas modulated by attention correspond closely to those
showing activation during passive visual stimulation (Martinez et al., 2001), and that attention
to a particular attribute of a visual stimulus (e.g. color, orientation, motion) enhances activity
in the visual area specialised for processing the selected attribute (Corbetta et al., 1991). We
suppose that attention influences neuronal firing in those cortical areas that are anatomically
recruited by attended stimulus because only those cortico-striatal inputs could be modified
that are active during dopamine release (Silkis, 2000), and because this modification is
necessary for attentional filtering.
The real stimulus with expected properties should cause strong initial cortical reaction
due to summing up real excitation with anticipating activity. Since in this case cortical
response is strong the neuronal pattern could be further contrasty selected by the C-BG-Th-C
loops. For the same reason visual attention to a stimulus feature could facilitate the processing
of other stimuli sharing the same feature. Such effect is often obtained (Saenz et al., 2003).
According to our model, the attention directed on a certain property of stimulus strengthens
responses of those cortical neurons for which this property is preferable because their
reactions are initially large and cortico-striatal input is strong. Simultaneously attention
suppresses responses of neurons for which this stimulus property is not preferable since their
responses are initially poor and cortico-striatal input is weak.
Remarkably, the earliest component of visual responses enhanced by attention was
obtained in the extrastriate cortex in the time range of 80-130 ms after stimulus onset
(Hillyard and Anllo-Vento, 1998; Martinez et al., 2001), whereas neuronal responses with
latencies of 20-30 ms and 50-55 ms were not influenced by attention (Anllo-Vento et al.,
1998; Di Russo et al., 2003; Maunsell and Gibson, 1992; Vidyasagar, 1998). If the attention
is based only on recurrent cortico-cortical and/or cortico-thalamic projections as it is
commonly proposed (Woldorff et al., 2002), the short latency components of responses
should be also amplified since time lags of mentioned connections are small. Our model
explains these results by necessity of activation of dopaminergic cell for attentional effects.
By this reason, attention can increase only those components of reactions in different cortical
areas whose onset exceeds the latency of visual responses of dopaminergic cells, which is
about 100 ms (Dommett et al., 2005; Schultz et al., 1997).
It was found that identification of the second of two targets is impaired if it is presented
less than about 500 ms after the first (Di Lollo et al., 2005). It was assumed that this effect,
known as attentional blink, is more probably the result of temporary loss of control over the
prevailing attentional set (Di Lollo et al., 2005). From point of view of our model, attentional
blink could be explained by temporal characteristics of dopaminergic cell responses and
release of dopamine in the striatum. Light flashes cause release of dopamine in the striatum
with the mean latency 154 ms and mean duration about 331 ms (Domett et al., 2005). Since
increase of excitation of dopaminergic cells in response to a light flash is followed by a
decrease of firing rate that lasts about 150 ms (Domett et al., 2005), more than 500 ms is
necessary for the normal response of dopaminergic cell to the second visual stimulus (i.e. for
